KR20000012012A - luminescent substrate LED device - Google Patents

Abstract

PURPOSE: A LED is provided to radiate an intermediate color such as pink, red, yellow and white and so on by a single chip. CONSTITUTION: The radiation LED device has; a substrate made from a semiconductor crystal that adds impurities as a fluorescent center or an insulating crystal; and an epitaxial layer formed on the substrate and having a structure that is radiated by injecting a current. Then the device changes color of the device radiation by changing the structure element composition, the impurity element sort and the substrate thickness of the crystal substrate and also changing the structure of the epitaxial layer.

Description

Luminescent substrate LED device

The present invention relates to a structure of a novel semiconductor light emitting device having two kinds of light emitting peaks as a single device. More specifically, the present invention relates to a single LED structure in which a light emitting layer is epitaxially grown on a substrate, wherein the light of the light emitting layer and the substrate fluorescence are synthesized to generate intermediate colors such as pink, red, orange, yellow, and white.

High-brightness light emitting diodes (LEDs) generating various monochromatic light have already been put into practical use. As a red LED, LED which uses AlGaAs, GaAsP, etc. as a light emitting layer is used widely. High brightness of several Cd (candelas) or more is produced in low cost. Several monochromatic LEDs other than red have also been put to practical use. LEDs using green / yellow green GaP, blue SiC, blue / green GaInN, and orange / yellow AlGaInP as active layers have been put into practical use as low cost LEDs. The substrate is GaAs, GaP, SiC, sapphire or the like.

The relationship between the color of the LED which is put into practical use and the composition of the active layer is as follows.

(1) Red LED ... AlGaAs, GaAsP

(2) Green / Yellow Green LED ... GaP

(3) Blue ... SiC

(4) Blue, green ... GaInN

(5) Orange, yellow ... AlGaInP

However, all of these LEDs use light emission due to this interbandgap transition in the active layer. The wavelength is determined to be one because of the band gap transition of electrons. It is monochromatic because it is a single wavelength of light. Since light emission is caused by recombination of electrons and holes in a band gap, it has a narrow spectrum. Since both are bandgap emission of a single semiconductor material, only a single color is produced.

In addition to the primary colors, several neutral colors may be produced. However, even a medium color does not change to a single color. The colors that can be generated by the LEDs so far are red, orange, yellow, yellow green, green, blue green, blue, blue purple, purple and the like. These are primary colors or intermediate colors of red and green or intermediate colors of green and blue. Even medium colors are solid colors, not complex colors. In addition, none of the intermediate colors of red and blue and the intermediate colors of red, green and blue were emitted by a single LED.

In order to produce intermediate colors, a combination of a plurality of red, green, and blue three-color LEDs is used. Combining the three primary LEDs can produce any intermediate color. However, combining several different LEDs complicates the structure. There is also a need for research in which color is invisible. It is an object of the present invention to impart intermediate color by a single LED rather than a combination.

Instead of the above monochromatic light source, a light source of red and blue intermediate colors (fuchsia and pink) and a light source of red, blue and green intermediate (white) are obtained for the purpose of illumination and some display purposes. As described above, only a single color comes out of the normal LED. Lighting and display require neutral colors. Therefore, lighting and display are also used as a daylight or incandescent lamp.

These lighting indicators also have advantages such as track record and easy handling. Moreover, the apparatus and the tube are inexpensive. There is an advantage of mature technology, such as being able to directly connect a commercial power supply. Incandescent bulbs, however, have a short lifespan. It is frequently cut off and needs to be taken frequently. Incandescent bulbs also have the disadvantage of poor luminous efficiency. Fluorescent lamps are improving for life, but have the drawback of being heavy with large devices.

LEDs are efficient because they convert current directly into light. Another long life. In addition, the individual devices are extremely light. Although LED is also used for the display purpose etc. which are good as a primary color, use is limited because it cannot produce colors, such as white, a reddish purple, and pink. The only way to achieve a neutral color is to combine several different LEDs. But it's a complex structure, making it a high cost.

Attempts have been made to synthesize white by a single LED. This is a combination of a high brightness blue LED using GaInN and a YAG yellow phosphor. Since a technology for producing a blue LED by growing a GaN crystal on a sapphire substrate and growing a GaInN active layer has been established, the blue LED is melted. This white LED is introduced in Shuji Nakamura & Gerhard Fasol, "The Blue Laser Diode (GaN Based Light Emitters and Lasers)", January 1997, springer, p 216-221 (1997).

1 shows a known white LED indicated in the document. A blue LED 5 having a GaInN / GaN structure formed on a sapphire substrate is bonded in the recess 4 of the stem 2. The P side (anode) electrode and the n side (cathode) electrode of the LED are on the upper surface of the device, but they are wire-bonded to the stems (2) and (3). The YAG phosphor 6 is filled on the GaInN-LED 5. YAG (yttrium aluminum garnet) is a yellow material and absorbs blue light to generate yellow fluorescence.

Conventional light emitting devices or light receiving devices use a conductive substrate, so the bottom of the chip becomes an electrode and is fixed to the lead. Therefore, the wire also ends with one connecting the upper electrode on the side and the lead. However, GaN-based blue LEDs stack a GaN or GaInN layer on a sapphire substrate. Since sapphire is an insulator, the bottom cannot be a cathode. Therefore, the n electrode (cathode) and the p electrode (anode) are arranged in parallel on the upper surface of the chip. Therefore, two wires are needed. When a current flows from the anode to the cathode, the GaN LED emits blue color. Part of the blue light passes through the YAG phosphor as it is and is emitted to the outside. The rest is absorbed by the phosphor 6 to give a longer wavelength yellow. Blue and yellow light overlaps. The synthesized light is white. This, in turn, means that the blue of the GaN-based LED and the yellow from the phosphors excited by it are synthesized, which is perceived by the human eye and thus appears white.

The light emission of LED is active light emission by the interband transition of the former. The phosphor absorbs the light and generates light when the electrons inside are excited from the base band to the upper band and fall to the base band via a level called the light emitting center. Naturally, the excitation light emits light having a lower energy than that of the LED. Surrounding the LED with a suitable phosphor produces the inherent light of the LED and fluorescence of longer wavelengths. The YAG phosphor emits yellow light, so it is synthesized with the blue color of the LED and becomes white. In visible light, blue has a short wavelength and high energy. Such a thing becomes possible because a blue light emitting element exists.

Fig. 2 shows light emission spectra of the GaInN / YAG light emitting device. The horizontal axis represents wavelength, and the vertical axis represents light intensity (arbitrary scale). There is a sharp emission peak around 460nm. This is a peak inherent in blue LEDs (GaInN-LEDs). There is a broad peak centered at 550 nm, but this is light emission by fluorescence of yellow YAG. It emits blue and yellow by electron transition emission and fluorescence. The human eye cannot see the color separately, so it looks like white light. It means that even one LED can make white.

However, there are some drawbacks to the GaInN / YAG light emitting device. GAInN-based LEDs require extra YAG phosphors that are completely heterogeneous. This is the first drawback. A large amount of light from the LED is absorbed by filling the YAG phosphor with poor transparency on the chip. The blue color of the LED does not come out to the outside so that the efficiency is extremely bad. The blue LED alone has a brightness of 1 Cd (candela) or more and an external quantum efficiency of 5% or more, but when the white LED is surrounded by YAG, the brightness decreases to 0.5 Cd and the external quantum efficiency to 3.5%. This is a second drawback.

In addition, YAG fluorescent material is not only poor in transparency, but also has low light conversion efficiency and ability to make color. Only about 10%. It is necessary to increase the thickness of the fluorescent material when trying to raise the intensity of yellow light emission in order to make the color tone of the white light of the eye seen white to a warm color system. Doing so increases the absorption gradually and decreases the luminance and efficiency. This is a third drawback. The fourth drawback is that a complicated manufacturing process is required.

There is a great need for white LEDs that can be used for lighting, display, liquid crystal backlights, etc., but there is also a need for intermediate tones other than white. Yellow and orange, which are halfway between red and green, are expected to be in high demand for warning lights and display purposes. Single LEDs with orange or amber, half-red and green colors can be manufactured. But everything is low brightness and high cost. For example, an LED having a layer of a P-type Al GaInP layer on an n-type GaAs substrate by MOCVD produces yellow color. However, since the current expansion in the p-type layer is bad, AlGaInP must be thickened. Therefore, it becomes a high cost. In addition, the brightness is low. High cost, low brightness is a big challenge for LEDs. Instead, a combination of two LEDs, a more inexpensive red LED (AlGaAs) and a green LED (GaP), is used. In addition, research is also said to put on the cover glass painted. The combination of the two LEDs complicates the structure. The power is also not simple. Higher brightness single LEDs which are more inexpensive and give off orange and yellow colors are preferred. In addition, it has been impossible to produce a red blue mid color (pink, red purple) and a red green blue mid color by a conventional single LED. It is an object of the present invention to overcome the difficulties of the prior art and to provide LEDs which generate red, blue intermediate, white (red, blue, green intermediate) by a single LED.

A general chromaticity diagram is explained with reference to FIG. The chromaticity diagram shows the stimulus values for the three primary colors of red, green, and blue (the amount of stimulation felt by three types of visual senses called `` gatherings '' in the human eye) with respect to the general visible light source color or object color. It is a figure studied to display on a plane coordinate by digitizing. When the light emission spectra of an arbitrary light source is Q (λ), the value of the stimulus of each color is multiplied by the equivalent color function corresponding to the spectral sensitivity characteristic of the visual sensor that recognizes each color. As a result, if the orange color function corresponding to red is r (λ), the green color function corresponding to green is g (λ), and the orange color function corresponding to blue is b (λ), the red stimulus amount X is X = ∫Q. (λ) r (λ) dλ, green stimulus amount Y is Y = ∫Q (λ) g (λ) dλ, and blue stimulus amount Z is Z = ∫Q (λ) b (λ) dλ. The planar coordinates unfolded by x = X / (X + Y + Z) and y = Y / (X + Y + Z), the lower limit of which is the total stimulus amount, are chromaticity diagrams shown in FIG. 17. In principle, in this coordinate system, the chromaticity of any spectra is represented as a point inside a rectangular isosceles triangle that can be connected by connecting (0,0), (1,0), (0,1) on the coordinates. .

In the chromaticity diagram, monochromatic light is represented by a solid C-shaped solid line in FIG. Since this type is determined from the shape of the orange color function, for example, in the wavelength region longer than 550 mm, since the blue sensitivity is 0, the chromaticity of the monochromatic light exists on a straight line of x + y = 1. In addition, in the wavelength region shorter than 505 nm, blue increases and the spectral sensitivity corresponding to red only increases. Therefore, a curve deviated from the straight line x = 0 (y-axis) is drawn as shown in the drawing. Although the extreme point of the long wavelength and the short wavelength of the C-shaped curve are connected by a straight line, the straight line does not correspond to monochromatic light and is called a pure purple trajectory. The inner point of the enclosed area between the C-shaped curve and the pure purple color trajectory displays the intermediate color. The center of this intermediate color is the white area. As can be seen from FIG. 19, the white region exists in the range of x = 0.22 to 0.49 and y = 0.2 to 0.46. The color aimed at by this invention was not possible with the conventional LED. The white area, and the pink, purple, reddish-purple mid-range area below it, and the yellow and orange red-green intermediate area (the upper right side of a white area) which had only the high cost low luminance product in the conventional LED.

Fig. 1 (a) is a sectional view of the entire white LED device according to the conventional example in which the GaInN-LED is buried in a YAG fluorescent material.

Fig. 1 (b) is a sectional view of the vicinity of the YAG fluorescent material and the LED chip of the white LED device according to the prior art, which embeds the GaInN-LED on the YAG fluorescent material.

FIG. 2 is a light emission spectrum diagram of the white LED of FIG.

Figure 3 (a) is a cross-sectional view of the whole intermediate, white LED device according to an embodiment of the present invention

Fig. 3 (b) is a sectional view of only the vicinity of the chip of the intermediate and white LED device according to the embodiment of the present invention.

Figure 4 is a spectrogram for explaining the principle of the LED of the present invention using the current injection light emission and the substrate fluorescent light emission. Wavelength A is the longest wavelength for generating fluorescence, wavelength B is the peak center wavelength of fluorescence, and wavelength C is the peak center wavelength of band gap transitional emission by current injection.

5 is a layer structure diagram of a light emitting substrate LED according to Embodiment 1 of the present invention (ZnSe.ZnCdSe / ZnSe (I.Cu)).

Fig. 6 is a light emission spectrum diagram of a light emitting substrate LED according to Embodiment 1 of the present invention (ZnSe.ZnCdSe / ZnSe (I.Cu)).

Fig. 7 is a chromaticity diagram showing the emission chromaticity of a light emitting substrate LED according to Embodiment 1 of the present invention (ZnSe.ZnCdSe / ZnSe (I.Cu)).

8 is a layer structure diagram of a light emitting substrate LED according to Embodiment 2 of the present invention (ZnSe / ZnSe (Al · I)).

Fig. 9 is a light emission spectrum diagram of a light emitting substrate LED according to Embodiment 2 of the present invention (ZnSe / ZnSe (Al · I)).

Fig. 10 is a chromaticity diagram showing the emission chromaticity of a light emitting substrate LED according to Example 2 (ZnSe / ZnSe (Al · I)) of the present invention.

Fig. 11 is a layer structure diagram of a light emitting substrate LED according to Embodiment 3 of the present invention (ZnSSe.ZnCdSe / AlGaAs (Si)).

Fig. 12 is a light emission spectrum diagram of a light emitting substrate LED according to Example 3 (ZnSSe.ZnCdSe / AlGaAs (Si)) of the present invention.

Fig. 13 is a chromaticity diagram showing emission chromaticity of a light emitting substrate LED according to Example 3 (ZnSSe.ZnCdSe / AlGaAs (Si)) of the present invention.

Fig. 14 is a layer structure diagram of a light emitting substrate LED according to Embodiment 4 (GaInN / GaP (Zn · O)) of the present invention.

Fig. 15 is a light emission spectrum diagram of a light emitting substrate LED according to Embodiment 4 (GaInN / GaP (Zn · O)) of the present invention.

Fig. 16 is a chromaticity diagram showing the emission chromaticity of the light emitting substrate LED according to Example 4 (GaInN / GaP (Zn · O)) of the present invention.

Fig. 17 is a chromaticity diagram showing the light emission chromaticity of all the LEDs (α to η, κ to μ) of Examples 1 to 6 collectively;

Fig. 18 shows the materials, fluorescence wavelengths, active layer materials, current injection light emission wavelengths, LED symbols, upper points (substrate thickness, impurity concentration, active layer) of the substrates of all the LEDs (?-?,?-?) Of Examples 1-6. Table showing the chromaticity of the LEDs

Fig. 19 is a general chromaticity diagram showing an intermediate color region generated by expressing a monochromatic color by a contour line with a red component on the x-axis and a green component on the y-axis.

Fig. 20 shows the layer structure of the epitaxial wafer of the LED of Example 5 having GaInN as an active layer;

Fig. 21 is a light emission spectrum diagram of an LED used in the present invention having GaInN as an active layer.

Fig. 22 is a chromaticity coordinate system in which the chromaticity coordinates of which the horizontal axis is X and the vertical axis are Y, in which light emission points of LEDs having three different thicknesses of Example 6 are represented as κ, λ, μ. M represents light emission of GaInN-LED, and N represents fluorescence. X represents the ratio of the red component of three primary colors, and Y represents the ratio of the green component of three primary colors.

<Description of the symbols for the main parts of the drawings>

1, 15: transparent resin 2,3,11: stem

4: recess of stem (2) 5: GaInN-LED

6: YAG phosphor 7,8,14: wire

9: LED chip 10: lead

12: substrate 13: epitaxial light emitting structure

15: transparent resin 16: element stand

20: n-type ZnSe substrate 21: n-type ZnSe buffer layer

22: n-type ZnMgSSe cladding layer

23: ZnSe / ZnCdSe heavy weight well active layer

24: p-type ZnMgSSe cladding layer 25: p-type contact layer

30: n-type ZnSe substrate (I, Al dope) 31: n-type ZnSe buffer layer

32: n-type ZnMgSSe cladding layer 33: ZnSe double hetero active layer

34: p-type BeMgZnSe cladding layer

35: p-type ZnTe / ZnSe superlattice contact layer 40: n-type AlGaAs substrate

41: n-type ZnSSe buffer layer 42: n-type ZnMgSSe cladding layer

43: ZnSSe / ZnCdSe heavy weight well active layer 44: p-type ZnMgSSe cladding layer

45: p-type ZnTe / ZnSe superlattice contact layer 50: semi-insulating GaP substrate (Zn, O-doped)

51: n-type GaN contact layer 52: n-type AlGaN cladding layer

53: GaInN double hetero active layer 54: p-type AlGaN cladding layer

55: p-type GaN contact layer 60: GaN epitaxial wafer

62: n-type GaN substrate 63: n-type GaN buffer layer

64: n-type AlGaN cladding layer 65: GaInN active layer

66: p-type AlGaN cladding layer 67: p-type GaN contact layer

The present invention does not use a fluorescent material, impurity is caused to emit light by the semiconductor substrate or the insulator substrate itself, thereby giving an LED which generates white or other intermediate colors in combination with light emission by a band gap transition. do. The epitaxial growth layer (light emitting structure) on the substrate emits a short wavelength of strong light by the band gap transition, and the substrate itself causes the fluorescence of a longer wavelength to be generated. After all, it uses LED light emission and fluorescence of substrate.

The LED is composed of a single crystal substrate and a light emitting structure epitaxially grown thereon. When the wavelength of the light emitted by the epitaxially grown layer is λ ₁ , the substrate must be normally transparent to this. If the substrate is opaque and absorbent, research has been done to remove it. If color impairs vividness due to impurities, studies have been made to minimize the impurities in the substrate. The substrate should be as transparent as possible.

In contrast, the present invention actively imparts a light emitting center to the substrate. Adding a suitable dopant to the substrate results in a light emission center. Doping imparts fluorescence to the substrate. Neutral and white are synthesized by a combination of fluorescence and LED emission. There are several combinations of LED substrates and active layers. For example, the following four combinations are possible.

Combination (a) GAN substrate GaInN active layer

Combination (b) AlGaAs substrate

ZnSe, ZnSSe / ZnCdSe, ZnSeTe Active Layer

Combination (c) GaPN GaInN active layer

Combination (d) ZnSe substrate

ZnSe, ZnSe / ZnCdSe, ZnSeTe Active Layers Among these substrates, the GaP substrate of the combination (c) can be easily prepared. A large single crystal is made by the chocolate skiing method or the bridging method, which is cut into pieces to form a single crystal wafer. Since the AlGaAs substrate of the combination (b) cannot form a large single crystal, an AlGaAs layer is formed on the GaAs wafer to form a substrate. The GaN substrate itself of the combination (a) is new and its manufacturing method is not widely known. The manufacturing method of the GaN substrate is also described later. The ZnSe substrate in combination (d) is also not well known. Therefore, this manufacturing method will be described later. In order to make it a fluorescence center, the impurity doped to each board | substrate is as follows.

Combination (a) GaN substrate Oxygen atom, Carbon atom, Nitrogen empty hole

Combination (b) AlGaAs substrate Si

Combination (c) GaP substrate Zinc atom, Oxygen atom

Combination (d) ZnSe substrate I + Cu, I + Ag, I + Al

As described above, the dopant type, concentration, and defect density can be adjusted to change the central wavelength of fluorescence emission. The ratio of fluorescence can be changed by the thickness of the substrate. Therefore, by arbitrarily adjusting parameters such as impurity type, concentration, defect density, substrate thickness, and the like, any white or various intermediate colors can be obtained from a warm color system to a white color system.

When λ ₁ (center wavelength C) emerges due to a bandgap transition in the active layer, the fluorescence center of the substrate absorbs it and emits fluorescence of λ ₂ (center wavelength B) having a low energy (λ ₁ <λ). ₂ ). This produces light of λ ₁ (center C) + λ ₂ (center B), so that the human eye is seen as their synthetic color.

Changing the material of the active layer changes the wavelength λ ₁ of the light of the band gap transition. Λ ₂ can be changed by changing the type of dopant in the substrate. Changing the amount of dopant or changing the thickness of the substrate changes the intensity of fluorescence. Therefore, by changing the epitaxial light emitting structure and the dopant of the substrate, various intermediate colors and whites can be generated. Since conventional LEDs necessarily have a substrate and a light emitting structure, the present invention does not need to add other structures. It only increases the process of doping the substrate.

4 shows the principle of the LED of the present invention. Wavelength C indicates light emission due to current excitation. Wavelength B is substrate fluorescence. To understand the present invention, it is necessary to understand the relationship between the three wavelengths A, B, and C.

Wavelength C = light emission due to band gap transition by epitaxial light emission structure

Wavelength B = fluorescence on the substrate

Wavelength A = the longest wavelength that can be absorbed by the substrate and fluoresce

Since the light having the lowest energy that can be broken by the fluorescent light B is A, in the case of one-photon excitation, the wavelength B must be wavelength A. When the light C emitted by the light emitting structure is wavelength C <wavelength A, the light C of the light emitting structure can excite fluorescence in the substrate. The band gap transition light C and the fluorescent light B are emitted to the outside. Light coming out is (C + B). The wavelength A is only a wavelength having a critical meaning and is not external light.

3 (a) and 3 (b) show a conceptual diagram of the LED of the present invention. The LED chip 9 of the present invention is mounted on the table 16. The LED chip 9 is composed of a substrate 12 and an epitaxial light emitting structure 13. When there is an electrode on the back surface of the substrate, the electrode is connected to the stem 16. Moreover, one electrode is connected to the stem 11 by the wire 14. The entire chip and the steps above are molded by the transparent resin 15.

As the dopant of the substrate 12, one that absorbs light having a shorter wavelength than wavelength A of FIG. 4 and emits fluorescence having a peak at wavelength B (λ ₂ ) is selected. The epitaxial light emitting structure 13 has a wavelength C of band gap light emission reaching λ ₁ . On the other hand, A is the wavelength of the light of the lowest energy that can excite fluorescence. Light having a wavelength below the wavelength A may generate fluorescence. When the wavelength C emitted by the epitaxial light emitting structure is C <A, the light can excite the dopant of the substrate to generate fluorescence. For this reason, C <A <B may be any combination of substrates and epitaxial light emitting structures. However, the material obtained as a large crystal substrate is limited. Because of lattice matching conditions, thin films that can epitaxially grow on any substrate are also limited.

In the present invention, since light emitted from the substrate is fluorescence, it is natural that the energy of incident photons is lower than that of incident photons in the case of single photon absorption, and the fluorescence wavelength λ ₂ is longer than the excitation light λ ₁ . In the case of two-photon absorption, the fluorescent side may have a short wavelength. In the case where the light emission from the epitaxial light emitting structure or the fluorescence from the substrate is significantly shifted from the visible light region, the object of the present invention is not met.

In Fig. 4, the sharp peak centered on the wavelength C is light emission due to the band gap transition of electrons from the epitaxial light emitting structure. Broader spectr B with a longer wavelength than that indicates the fluorescence of the substrate. Light having a wavelength shorter than the wavelength A may excite the emission center of the substrate to fluoresce. The wavelength of the light emitted from the active layer is shorter than the wavelength A. Light of a short wavelength is emitted from the active layer, which excites the emission center of the substrate to emit long wavelength fluorescence. The synthesized light is emitted to the outside, but this is a combination of bandgap emission C and fluorescence B. By changing the epitaxial emission structure, the wavelength of the bandgap emission C can be changed. After all, changing the active layer structure can change the wavelength C. The fluorescence wavelength B can be changed by changing the dopant of the substrate. Changing the substrate dopant concentration can increase or decrease the ratio of fluorescence. By changing the substrate thickness, the ratio of fluorescence can be added or subtracted.

As a result, fluorescence (substrate emission) having a wavelength B as a peak has parameters that can be changed freely such as dopant, dopant concentration and thickness. In the epitaxial light emitting structure, the active layer can be parameterized to fluctuate wavelength C freely.

The present invention is to change the epi-luminescence C, fluorescence B, it is possible to obtain an LED that generates a wide range of intermediate colors, white. The LED of the present invention can also generate a red-blue medium color and a white (red-blue color medium), which have never been possible with LEDs. It is indeed an excellent invention.

The shape of the LED chip of the present invention is not changed from the conventional LED chip. There is no need to apply extra fluorescent material. Stems can also use common ones. The already established low cost LED device manufacturing procedure can be applied as it is.

In general, LEDs require a substrate as a stand for forming an epitaxial layer. The substrate is necessary. In the conventional LED, the substrate has no active role other than keeping the epi layer to pass current. In the case where the substrate emits light, studies have been conducted to remove substrate emission by removing impurities. It has been suggested that the substrate is transparent, absorptive and not emitting light. In the present invention, substrate fluorescence is actively used in reverse. It is a technology created by the reversal of the idea of trying to produce neutral or white color using fluorescence.

In the present invention, Japanese Patent Application Laid-open No. Hei 10-194156 proposes a technique for producing a white light LED using a ZnSe substrate + ZnSe epitaxial layer by combining bandgap transition light emission and fluorescence. This is synthesized by generating blue in the ZnSe-based active layer and yellow in the ZnSe substrate to produce white light. However, by studying the dopant of the substrate, the type of substrate, and the epi layer, a wide range of intermediate colors can be created. Red-blue, mid-blue, and red-green mid-colors, which have not been possible until now, can also be produced by LEDs. The reddish green color was also due to the conventional LED but was high cost. The present invention can generate a reddish green intermediate color by an inexpensive LED.

For example, the present invention proposes a white LED having a simple structure which only combines a GaN substrate including a fluorescent center with a GaInN-based blue light emitting device (combination (a)). Growing a p-type GaInN thin film on an n-type GaN substrate results in a blue LED emitting light at 400 nm to 500 nm.

GaN substrates with large size and few defects cannot be manufactured conventionally. GaN itself is not easily melted even at high temperature and high pressure, and cannot be used with the chocolate skiing method or the bridging method. Thus, GaInN-LED is conventionally formed on a sapphire substrate. The present invention cannot be implemented unless a GaN substrate is present. In recent years, GaN substrates can be produced by the melt growth method or the vapor phase growth method. This made the present invention possible. The melt growth method is a method in which a single crystal of GaN is grown by dissolving GaN in a Ga melt and applying a pressure and heat to form a melt of Ga-GaN. Can make small crystals.

In the vapor phase growth method, a mask having a plurality of dot-shaped holes or linear holes is formed on a (111) GaAs substrate, and GaN is vapor-grown at low temperature through a mask to form a buffer layer, and an epitaxial layer at high temperature thereon. Is grown thick and GaAs substrate is removed to make GaN large single crystal substrate. In the end, the substrate was grown using the method of growing a thin film. Examples of the source gas supply method include the HVPE (helide vapor phase growth method) method, the MOC (organic metal chloride vapor phase growth method) method, and the MOCVD (organic metal CVD method) method. In the HVPE method, GaN is synthesized by placing GaAs substrate and Ga metal solution in a hotwall furnace, exhaling hydrogen and HCl gas to make GaCl, and reacting with ammonia near the substrate. MOC reacts an organic metal such as TMG with H ₂ + HCl gas and a hot wall furnace to synthesize GaCl, and reacts this with ammonia NH ₃ to form GaN. In a cold wall reactor, MOCVD transports organic metals such as TMG by H ₂ , and reacts with ammonia to form GaN. The method for producing these GaN substrates is described in Japanese Patent Application Laid-Open No. 10-171276, which is the source of the applicant. It is extremely recent that it is possible to manufacture large GaN substrates. The present invention produces a white LED using such a GaN substrate as a starting material.

When the GaN substrate is grown by the vapor phase growth method or the melt growth method, impurities (dopants) such as oxygen and carbon can be doped or crystal defects (nitrogen void holes) can be introduced. Impurities such as oxygen, carbon, or crystal defects such as nitrogen pores become a center for generating fluorescence. When light of a wavelength shorter than 480 nm is touched, a wide range of fluorescence of 520 mm to 650 mm is generated. This light emitting center may be referred to as a fluorescent light emitting center or simply a fluorescent center. The central wavelength of the fluorescence emission and the half width of the emission spectra can be adjusted by the type of dopant (oxygen and carbon), the doping amount, or the amount of crystal defect (nitrogen void hole). Since fluorescence is widely distributed from yellow to red (520 nm to 650 nm), this and the blue of GaInN-LED are added. It is synthesized by the naked eye, so it appears to be white or neutral light. White light or neutral light is synthesized from the two lights, such as white light, neutral light = blue light of GaInN and fluorescence of GaN.

As a result, the device of the combination (a) of the present invention consists of two parts,

(1) GaInN-based LED blue emission due to band transition (400-510 nm)

(2) GaN substrate ... A combination of yellow to red fluorescence (fluorescence; 520 to 650 nm). The GaN substrate may be n type or p type. In either case, a Pn junction is formed in the epitaxial growth layer. The epitaxial growth layer becomes a light emitting structure. The substrate becomes a phosphor.

There are various things even if it is white. If blue predominates, it becomes a white color. If red predominates, it tends to warm color. If the GaN substrate is thick, the blue color of GaIn-LED is absorbed and reduced, and the yellow color of fluorescence is dominant. By changing the thickness of the GaN substrate, the intensity of fluorescence emission can be adjusted. Finally, the substrate thickness can change the ratio of fluorescence emission to blue emission from the LED. However, the substrate thickness was limited from other conditions. If it is 50 micrometers or less, the rate of breakage will increase in a post process. Yields drop and costs go up. On the contrary, when the substrate thickness is 2 mm or more, the size of the LED becomes too large. In addition, the ratio of the yellow light is excessively increased to become white. Therefore, the substrate thickness is about 50 to 2 mm. The same can be said for other combinations (b) to (d).

In the above, the manufacturing method of the GaN board | substrate of combination (a) was demonstrated. The substrate AlGaAs / GaAs of the combination (b) and the substrate GaP substrate of the combination (c) will not be described since the manufacturing method is well known. The manufacturing method of the n-type ZnSe substrate of the combination (d) is not necessarily known, and will be described briefly here.

As a method for producing a large ZnSe single crystal, there are a chemical transport method and a grain growth method. The chemical transport method transports a part of polycrystalline ZnSe by I and deposits it on monocrystalline ZnSe. ZnSe polycrystalline raw materials are placed at the bottom of the growth chamber. A single crystal ZnSe seed crystal is fixed to the growth chamber. Fill the space with oxo. The bottom polycrystal is heated to a higher temperature than the single crystal at the top. At the bottom, a reaction of 2ZnSe + 2I ₂ → 2ZnI ₂ + Se ₂ takes place. Zinc oxide ZnI ₂ rises because it is a gas. Se ₂ also rises. Seed crystals are low temperature, so the opposite reaction occurs here. 2ZnI ₂ + Se ₂ → 2ZnSe + 2I ₂ . This ZnSe is deposited by trimming the bearing on top of the seed crystal. Oxo serves to transport zinc in this way. It is also called oxo transport method. The growth temperature is about 850 ° C.

The grain boundary growth method locally heats ZnSe solid polycrystalline raw materials and moves heating elements to move the grains (grains) to promote integration between grains, reducing the number of grains and increasing the grain size. It is a method of manufacturing a single crystal by using only one crystal grain predominantly. The oxo transport method may produce a ZnSe single crystal containing oxo anyway, but the grain boundary growth method may produce one containing no oxo.

DETAILED DESCRIPTION Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

EXAMPLE

Example 1 (ZnSe substrates (I, Cu): ZnSe-based half-color elements; 480 nm, 630 nm)

As the substrate, oxo and copper-doped n-type ZnSe substrates were employed. As an epitaxial light emitting structure, a laminated structure composed of a mixed crystal having ZnSe as a mother was produced (FIG. 5). ZnSe substrate 20 doped with oxo and copper absorbs light having a wavelength shorter than 510 nm having a longer wavelength than 460 nm, which is a wavelength corresponding to the original band gap energy (wavelength A: 510 nm) Fluorescence (wavelength B: 630 nm) having a broad peak at 630 nm by (oxo and copper) is emitted. Band tailing phenomenon plays an important role in the present invention. The semiconductor of the bandgap Eg absorbs light having a wavelength λ <(λg) shorter than the wavelength λg (= hc / Eg) corresponding thereto and transmits light having a long wavelength λ (> λg). When impurity levels are generated at the end of the conduction band and the end of the valence band by doping the impurities, the transition of the impurity level and the valence band and the transition around the conduction band and impurities occur. As a result, light of a wavelength of λg or more can be absorbed. This is band tailing phenomenon. In this embodiment, the substrate can absorb light up to a wavelength A (510 nm) longer than lambda g (460 nm) by band tailing. As a result, impurity emission can be caused by light having a wavelength of up to wavelength A (510 nm).

In order to examine the influence of the thickness, the ZnSe substrate 20 prepared a substrate having a thickness of 50 μm (element α) and a substrate having a thickness of 200 μm (element β). An epitaxial luminescent structure as shown in Fig. 5 was grown on two kinds of thickness ZnSe substrates by the MBE method. The emission peak wavelength of this epitaxial light emitting structure is 480 nm (wavelength C: 480 nm).

The epitaxial light emitting structure is a p-type contact layer 25 consisting of a p-type ZnSe and a ZnTe laminated superlattice structure from the top, a p-type Zn _0.85 Mg _0.15 S _0.10 Se _0.90 cladding layer 24, ZnSe and ZnCdSe The multi-quantum well active layer 23 having a stacked structure of n-type Zn _0.85 Mg _0.15 S _0.10 Se _0.90 cladding layer 22 is formed. In fact, there is an n-type ZnSe buffer layer 21 between the ZnSe substrate 20 and the epitaxial light emitting structure. If we explain sequentially from the board side,

(1) n-type ZnSe substrate (I, Cu dope, impurity emission = 630 nm; B) 20

(2) n-type ZnSe buffer layer 21

(3) n-type Zn _.85 Mg _0.15 S _0.10 Se _0.90 cladding layer 22

(4) ZnSe / Zn _0.88 Cd _0.12 Se Heavyweight Well Active Layer (480 nm; C) 23

(5) p-type Zn _0.85 Mg _0.15 S _0.10 Se _0.90 cladding layer 24

(6) p-type ZnTe / ZnSe superlattice contact layer 25

It is the same layer structure. The active layer of (4) may be ZnSe _0.99 Te _0.01 double hetero active layer (480 nm: C).

On the p-type contact layer of the epi wafer, a dot-shaped pattern p electrode for each chip unit made of Pd / Au was formed, and a thin film Au electrode having a thickness of 20 nm or less was formed on the entire upper surface. If the thickness of Au is 20 nm or less, light is passed through to form a transparent electrode. On the lower surface of the substrate, an n-side electrode made of In was formed. The epi wafer after electrode formation was cut out of chips having dimensions of 250 µm x 250 µm. The chip was fixed to an element stand (stem) to make an LED.

In this LED, since the wavelength C (480 nm) is shorter than the wavelength A (510 nm), among the light emitted from the epitaxial luminescent structure, the incident light on the substrate is absorbed by the substrate and generates fluorescence of the wavelength B (630 nm). . So the light coming out is 480nm + 630nm.

This LED was measured by the constant current mode. Typical luminous intensity was 20㎃ and 2mW.

The emission spectra of this LED are shown in FIG. As a design, it can be seen that light emission from a light emitting structure having a sharp peak at 480 nm and impurity light emission of a ZnSe substrate having a broad peak at 630 nm are combined.

The element α having a thin substrate (50 μm) had a small substrate fluorescence intensity and had a blue color of pink. The thick substrate (200 mu m) beta had a large substrate fluorescence intensity and a reddish violet color with a reddish color. Thick substrates contain many impurity emission centers. Therefore, it is understood that the thicker the substrate, the greater the fluorescence.

Fig. 7 shows the emission spectra of the same LEDs on the chromaticity diagram. The device α having a substrate thickness of 50 μm was a pink color with a blue group having a chromaticity of (x, y) = (0.34, 0.21). The device β having a substrate thickness of 200 μm was a reddish violet color with a red color of (x, y) = (0.50, 0.24). In Fig. 7, the chromaticity (? Point) of the light from the epitaxial luminescent structure and the chromaticity (? Point) of the substrate fluorescence are also shown. The chromaticity of element (alpha), (beta) rides on the line segment which connects these two points. Similar experiments were carried out using ZnSe substrates doped with oxo and silver, yielding almost the same results.

[Example 2 (ZnSe (Al, I): ZnSe-based pink-reddish element; 465 nm, 600 nm)]

As the substrate, an n-type ZnSe substrate (dopant: oxo and aluminum) was selected. As an epitaxial light emitting structure, a laminated structure composed of a mixed crystal having ZnSe as a parent was adopted.

ZnSe substrates doped with oxo and aluminum absorb bands shorter than 510 nm, which are longer than 460 nm, which is the wavelength corresponding to the original band gap (wavelength A: 510 nm). Fluorescence with a broad peak at 600 nm through the center (wavelength B: 600 nm).

The fluorescence wavelength is changed by the dopant. In Example 2, since I and Al (oxo, aluminum) are dopants, the fluorescence center wavelength is 600 nm. The addition of Al is for shifting the fluorescence into long wavelengths.

Here, the ZnSe substrate prepared two types of doping amount n of 1 × 10 ¹⁷ cm ⁻³ (element γ) and 5 × 10 ¹⁸ cm ⁻³ (element δ). Any thickness is 250 micrometers. The purpose is to determine how SA emission changes by changing the dopant concentration.

On this conductive ZnSe substrate, a blue luminescent epitaxial luminescent structure having a peak wavelength of 465 nm was grown by homoepitaxial growth by the MBE method as shown in Fig. 8 (wavelength C: 465 nm). The epitaxial luminescent structure is a p-type cladding consisting of a p-type contact layer 35 consisting of a stacked superlattice of ZnSe and ZnTe doped in p-type, and Be _0.20 Mg _0.20 Zn _0.60 Se doped in p-type from above. A layer 34, a double heteroactive layer 33 made of ZnSe, and an n-type cladding layer 32 made of n-doped Zn _0.85 Mg _0.15 S _0.10 Se _0.90 . If we explain sequentially from the board side,

(1) n-type ZnSe substrate (I, Al dope) 30

(2) n-type ZnSe buffer layer 31

(3) n-type Zn _0.85 Mg _0.15 S _0.10 Se _0.90 cladding layer 32

(4) ZnSe double hetero active layer (465 nm) 33

(5) p-type Be _0.20 Mg _0.20 Zn _0.60 Se cladding layer 34

(6) p-type ZnTe / ZnSe superlattice contact layer 35

It is the same layer structure.

Using this epiwafer, an LED was produced in the same manner as in Example 1. Also in this LED, since the wavelength C (465 nm) is shorter than the wavelength A (510 nm), in the 465 nm light emitted from the epitaxial light emitting structure, the incident on the substrate is absorbed by the substrate, and the wavelength B (600 nm) Generates fluorescence.

This LED was measured in a constant current mode and the typical luminance was 20 Hz, which was 1 mW. High luminance yellow group of pink and blue group of reddish purple light were obtained.

The emission spectra of this LED are shown in FIG. As the design, it can be seen that light emission from the epitaxial light emission structure having a sharp peak at 465 nm (by band gap transition) and SA light emission from a ZnSe substrate having a peak broad at 600 nm are combined.

The element γ of the substrate having a small amount of doping (1 × 10 ¹⁷ cm ⁻³ ) has a small fluorescence intensity of the substrate, and the element fluorescence of the substrate having a large doping amount (5 × 10 ¹⁸ cm ⁻³ ) has a substrate fluorescence intensity It is large. It can be seen that the fluorescence is intensified in proportion to the amount of doping.

Fig. 10 shows the emission spectra on the chromaticity diagram. The (1 × 10 ¹⁷ cm ⁻³ ) element γ having a small doping amount is a reddish purple of a blue group whose chromaticity is (x, y) = (0.34, 0.19). The (5 x 10 ¹⁸ cm ^-3 ) element δ having a large doping amount is pink in a yellow group with chromaticity (x, y) = (0.50, 0.29). In Fig. 10, the emission chromaticity (? Point) from the epitaxial luminescence structure and the chromaticity (? Point) of only the substrate fluorescence are shown. The element γ and the element δ ride on the line segment connecting these two points.

Example 3 (AlGaAs substrate (Si); ZnSe-based yellow / orange element; 520 nm, 550 nm, 690 nm)

As the substrate, an AlGaAs substrate 40 in which an n-type AlGaAs layer 47 was formed by a liquid layer epitaxial method (LPE) on a n-type GaAs substrate 46 doped with Si was employed. As the epitaxial light emitting structure, a laminated structure composed of a mixed crystal based on ZnSe was employed.

The AlGaAs substrate 40 can change the band gap by changing the Al composition. When the band gap energy is expressed by the corresponding wavelength, it is possible to change the band gap energy in the range of 570 nm to 860 nm. Here, a composition of Al _0.50 Ga _0.50 As was selected as the substrate side composition. Si was adopted as an n-type impurity doped in the AlGaAs substrate 40. The AlGaAs substrate of the Al composition absorbs a wavelength shorter than 640 nm (wavelength A: 640 nm) and emits fluorescence having a broad peak at 690 nm by recombination center emission of dopant Si (wavelength B: 690 nm).

On the conductive AlGaAs substrate 40, a green light emitting structure (Fig. 11) having a light emission peak wavelength of 520 nm (element?) Or 550 nm (element?) Was heteroepitaxially grown by the MBE method. The epitaxial luminescent structure is sequentially formed from above, p-type contact layer 45 composed of p-type doped ZnTe and ZnSe laminated superlattice structure, p-type Zn _0.90 Mg _0.10 S _0.15 Se _0.15 Se _0.85 layer Multi-quantum well active layer consisting of a stacked cladding layer 44, a ZnS _0.06 Se _0.94 layer and a Zn _0.70 Cd _0.30 Se layer (ε), or a ZnS _0.06 Se _0.94 layer and a Zn _0.60 Cd _0.40 Se layer (ζ) (43), an n-type cladding layer 42 composed of Zn _0.90 Mg _0.10 S _0.15 Se _0.85 layer doped in n-type. In fact, there is an n-type ZnSSe buffer layer 41 between the n-type AlGaAs substrate 40 and the epitaxial light emitting structure.

Opening the layer structure from the substrate side

(1) n-type GaAs substrate 46

(2) n-type AlGaAs layer (Si dope, SA emission = 690 nm) 47

(3) n-type ZnSSe buffer layer 41

(4) n-type Zn _0.90 Mg _0.10 S _0.15 Se _0.85 Clad Layer 42

(5) ZnS _0.06 Se _0.94 / Zn _0.70 (Cd _0.30 Se heavy weight well active layer (520 nm) 43 ... (element ε), or

(5 ') ZnS _0.06 Se _0.94 / Zn _0.60 (Cd _0.40 Se) Heavy weight well active layer (550 nm) 43 ... (element ζ)

(6) p-type Zn _0.90 Mg _0.10 S _0.15 Se _0.85 cladding layer 44

(7) The same lamination structure as the p-type ZnTe / ZnSe superlattice contact layer 45 is obtained. The active layer of the element ε of (5) may be ZnSe _0.90 Te _0.10 double hetero active layer (520 nm).

The elements ε and ζ have different mixing ratios of Cd of one layer ZnCdSe of the multi-quantum well active layer. The element ε is 0.30 and the element ζ is 0.40. The higher the Cd ratio, the narrower the band gap and the longer the emission wavelength.

For such an epi wafer, an LED was produced in the same manner as in Example 1. However, as the n-side electrode on the substrate side, Au-Ge electrode was adopted. The P side electrode is a Pd / Au electrode and is the same as in the previous example.

Although the LED and the epitaxial light emitting structure are different from the substrate, the light emitting wavelength (wavelength C) is shorter than the wavelength of the fluorescence (wavelength A). Part of the light (520 nm or 550 nm) emitted from the epitaxial light emitting structure is incident on the substrate and is absorbed by the substrate to generate fluorescence of wavelength B (690 nm).

This LED was measured by the constant current mode. High brightness yellow and orange light could be obtained. Typical luminance was 20 Hz and 3 mW.

The emission spectra of this LED are shown in FIG. As a design, it can be seen that a sharp light emitting structure (by band gap transition) having a peak at 520 nm (element ε) or 550 nm (element ζ) and impurity emission of an AlGaAs substrate having a broad peak at 690 nm are combined. .

Fig. 13 shows the emission spectra in the chromaticity diagram. The emission wavelength from the epitaxial light emitting structure is 520 nm (element?), And the chromaticity is yellow with (x, y) = (0.47, 0.48). The emission wavelength from the epitaxial light emitting structure is 550 nm (element ζ), and the chromaticity is orange with (x, y) = (0.57, 0.43). Also in Fig. 13, the chromaticity of light only from the light emitting structure (?: 2 types) and the chromaticity of the substrate fluorescent light (? Point) are shown. Together with element ε and element ζ, they are in the line segment connecting these points.

[Example 4 (GaP substrates (Zn, O); GaInN-based yellow elements; 520 nm, 700 nm)]

As the substrate, a semi-insulating GaP substrate 50 doped with Zn and O was employed. As an epitaxial luminescent structure, a mixed crystal structure of GaN as a matrix was adopted. GaP is an indirect transition type semiconductor, but absorbs wavelengths shorter than 550 nm (wavelength: 550 nm) by doping Zn and O, and emits fluorescence having a broad peak at 700 nm by recombination emission through an impurity recombination center ( Wavelength B: 700 nm).

On this semi-insulating (GaP substrate, an epitaxial luminescent structure as shown in Fig. 14 was heteroepitaxially grown by MOCVD. This is an epitaxial luminescent structure with a peak wavelength of 520 nm, which is good for green. )to be.

The epitaxial luminescent structure is a p-type contact layer 55 made of GaN doped in p-type, and a p-type cladding layer 54 made of Al _0.15 Ga _0.85 N-doped p-type, Ga _{0.70 from above.} Double hetero active layer 53 consisting of _0.30 N layers, n type cladding layer 52 composed of Al _0.15 Ga _0.85 N layer doped with n type, n type contact (buffer layer) layer composed of GaN doped with n type ( 51).

In the past, the buffer layer is present between the epitaxial light emitting structure and the GAP substrate 50. However, in this embodiment, the n-type GAN contact layer 51 serves as a buffer layer. Since the GAP substrate is insulative, the n-side electrode cannot be attached to the bottom. An epitaxial layer is cut out to a part of n-type buffer layer which consists of GaN, and n-side electrode is produced. Therefore, an n-type GaN buffer layer is expressed as an n-type GaN contact layer. The p-side electrode is Pd / Au, and the n-side electrode is Ti / Al. The p-side electrode is formed on the p-type contact layer, and the n-side electrode is formed on the partially exposed n-type contact layer.

If the epiwafer layer structure is recorded from below,

(1) Semi-insulating GaP substrate (Zn, O dope: 700nm) 50

(2) n-type GaN contact layer (with n-side electrode) 51

(3) n-type Al _0.15 Ga _0.85 N cladding layer 52

(4) Ga _0.70 In _0.30 N double hetero active layer (520 nm) 53

(5) p-type Al _0.15 Ga _0.85 N cladding layer 54

(6) p-type GaN contact layer 55

Becomes

On the epitaxial p-type GaN contact layer 55, a dot-shaped p-side electrode made of Pd / Au was formed at a cycle of 500 mu m x 500 mu m. This is the chip size. By the same period, a part of the epi layer was cut vertically by dry etching, and the n-type GaN contact layer 51 was exposed, and a dot-shaped n-side electrode made of Ti / Al was formed there. The n-side electrode is not in the bottom but is formed in the n-type contact layer on which the top surface is exposed. The epiwafer after electrode formation was cut out to the size of 500 micrometers x 500 micrometers, and the element was fixed to the mount stand (stem). Let this be the element (eta).

In the case of this element η, since the substrate is insulative, a conventional LED mounting method is not applied as shown in Figs. 3A and 3B. Like the sapphire / GaInN of Fig. 1 (a) and Fig. 1 (b), a wire is bonded from the upper side to the n-side electrode and the p-side electrode and connected to the lead.

The wavelength C (520 nm) is shorter than the wavelength A (550 nm) even with this LED element η. Therefore, among the light emitted from the epitaxial light emitting structure (520 nm), the incident light is absorbed by the GaP substrate and the wavelength B is used. Fluorescence of (700 nm). On the outside is a composite of 520 nm light and 700 nm light.

The LED was measured in a constant current mode and the light emission was yellow with a typical luminance of 20 Hz and 3 Hz. High brightness yellow light was obtained.

Fig. 15 is a graph showing emission spectra of this LED. As a design, it can be seen that light emission from an epitaxial light emitting structure having a peak at 520 nm (band gap transition) and impurity light emission from a GaP substrate having a broad peak at 700 nm are combined.

Fig. 16 shows this light emission spectrum on the chromaticity diagram. This light has yellow color with (x, y) = (0.45, 0.46). Also in FIG. 16, the chromaticity (Δ dot) for light emission from the epitaxial light emitting structure and the chromaticity (□ dot) of the fluorescence of the substrate are also shown. The chromaticity of this element lies on the line segment connecting the light emitting structure chromaticity with the substrate type chromaticity.

Example 5 (Gaseous Growth GaN Substrate, GaInN Active Layer, Grain)

The GaN single crystal is grown by vapor phase growth (HVPE, MOC, MOCVD method) or melt growth method as described above. The GaN self-supporting film is used as a substrate. GaN substrates impart defects such as dopants such as oxygen and carbon or nitrogen void holes. Oxygen, carbon and nitrogen pores become fluorescent centers. The thickness of the GaN substrate may be about 50 µm to 2000 µm. A GaInN crystal thin film is epitaxially grown on the GaN substrate. For example, epitaxial growth can be performed by MOCVD.

The structure of the GaN epitaxial wafer 60 is shown in FIG. An n-type GaN buffer layer 63, an n-type AlGaN cladding layer 64, a GaInN active layer 65, a p-type AlGaN cladding layer 66, and a p-type GaN contact layer 67 are formed on the n-type GaN substrate 62. do. In the epitaxial layer, Mg was used as the p-type dopant and Si was used as the n-type dopant. More specific composition is indicated.

(1) n-type GaN substrate 62

(2) n-type GaN buffer layer 63

(3) n-type Al _0.20 Ga _0.80 N Clad Layer 64

(4) Ga _0.88 In _0.12 N active layer 65

(5) p-type Al _0.20 Ga _0.80 N cladding layer 66

(6) p-type GaN contact layer 67

The contact layer needs ohmic contact with the p-electrode to have low resistance and high p-type concentration. The AlGaN cladding layer has a wide band gap and confines the carrier to the active layer. The GaInN active layer is obtained by alternately stacking a thin InN film and a GaN film on either layer.

A p electrode made of Pd / Au was formed on the p-type contact layer of the epitaxial wafer. An n electrode of In was formed in the n-type GaN on the back surface. In / TiAu may be used for the n electrode. Photolithography is used to form the pattern electrodes. The epitaxial wafer after electrode formation was cut out to the size of 300 micrometers x 300 micrometers angle, and it fixed to the stem 16 of the lid 10 like FIG.3 (a), FIG.3 (b). The n electrode was down and the p electrode was up. As a result, the GaN substrate 12 contacts the stem 16. The p-electrode was attached to the other lead 11 by a wire. These were molded by transparent resin.

This LED was measured by the constant current mode. High brightness white light was emitted. The typical luminance was 1.5 Cd for a drive current of 20 mA. Fig. 21 shows light emission spectra of this LED. By design, LED emission from an epitaxial luminescent layer with a sharp peak at 430 nm and fluorescence emission from a broad GaN substrate with a sharp peak at 550 nm are combined. The synthesized light is white. Blue was slightly strong white color.

Example 6 (GaN Substrate, Inverted, Substrate Thickness 3 Types of Melt Growth Method)]

Three kinds of GaN substrates having different thicknesses prepared using the melt growth method were prepared. Thickness is 100 micrometers, 500 micrometers, and 1 mm. On this GaN substrate, a homo epitaxial structure similar to Example 5 was produced by MOCVD. With the structure of FIG. 20, the p-type dopant is Mg and the n-type dopant is Si. In this structure, Zn is doped into the active layer.

(1) n-type GaN 62

(2) n-type GaN 63

(3) n-type Al _0.20 Ga _0.80 N Clad Layer 64

(4) Zn-doped Ga _0.88 In _0.12 N active layer 65

(5) p-type Al _0.20 Ga _0.80 N cladding layer 66

(6) p-type GaN contact layer 67

The p-electrode Pd / Au was provided on the p-type GaN contact layer. The p electrode is formed over the entire epitaxial surface. On the bottom of the n-type GaN substrate, an n-electrode of In having a thickness of 100 µm x 100 µm was provided. The chip coverage of the p-electrode is 100% and the chip coverage of the n-electrode is 11%. This was cut out with a square chip of 300 µm x 300 µm and mounted on the stem with the rear side facing up. It is also called episide down because the epitaxial surface is down. The p-electrode is mounted directly to the stem. Only the n electrode is connected to the stem by a wire. This was molded by transparent resin.

When this LED was made to emit light in the constant current mode, high brightness white light was obtained. The LED of Example 5 showed white tint stains, but this light receiving element did not have such spots in the element surface vertical direction. This is because the light emitted episide down and emitted in the vertical direction is a mixture of fluorescence and LED light emission. Typical luminance was a current of 20 mA and 1.5 to 2 Cd. Depending on the substrate thickness (100 μm, 500 μm, 1 mm), the white feel is wrong.

(κ) 100 μm-thick LED with a white tint of blue

(λ) 500 µm thick LED ... neutral white

(μ) LED having a thickness of 1000 µm. The chromaticities κ, λ, and μ of each white light emission are represented by chromaticity coordinates as shown in FIG.

In FIG. 22, point M indicates light emission from only the LED light emitting structure having a peak at 460 nm. Since color purity is good, chromaticity is also located at the corner of the curve. Point N represents the chromaticity of fluorescence emission centered at 580 nm. It is a broad peak and contains a variety of light, so it is inside the curve. The light which synthesize | combined the light of the point M and the light of the point N is on the straight line MN. When the GaN substrate is thin, the fluorescence weakens and is biased toward the M side. If the substrate is thick, the fluorescence becomes stronger and is biased toward N.

kappa ... GaN substrate thickness is 100 micrometers. It is closest to M point.

lambda point ... GaN substrate thickness is 500 mu m. In the middle.

µ point ... GaN substrate thickness is 1000 mu m (1 mm). It is closest to N point.

These white colors can also be expressed by color temperature.

κ point ... A little blue is a strong white color and the color temperature is about 80000K

λ point ... neutral white and bird's nest temperature is about 5000K

μ point ... yellowish yellow, warm white color, hue temperature about 3000K

By changing the thickness of the substrate in this way, a white color having a different color tone can be obtained. This is a case where the impurity concentration of the substrate is constant.

It was also found that changing the impurity concentration of the substrate changes the absorption coefficient of the substrate and the relationship between the substrate thickness and the color tone of the LED changes from the above relationship.

The emission wavelength can be changed between 40 nm and 500 nm by changing the composition X of Ga _1-X In _X N of the active layer. It has also been found that even if the LED light emission wavelength is changed in this range, white light can be obtained by optimizing the substrate thickness.

In Fig. 17, the chromaticity of the?-?,?-? LEDs described in Examples 1 to 6, and the chromaticity of their epitaxial luminescent structures and substrate fluorescence are collectively displayed.

Fig. 18 collectively displays the substrate materials of Examples 1 to 6, the fluorescence center wavelength, the material of the epitaxial light emitting structure, the wavelength of the current injection light emission, the substrate thickness, the symbol of the embodiment, and the like.

The present invention proposes an LED having an active layer suitable for using a GaN substrate, an AlGaAs substrate, a GaP substrate, a ZnSe substrate, and the like doped with impurities. Various intermediate colors and whites can be obtained by light emission of the active layer and substrate fluorescence. By the conventional single LED, it was not possible to produce a red total color neutral color and a red blue green intermediate color. According to the present invention, a single LED of red, blue, middle (red, pink), green, red, middle (yellow, orange), or red, blue, green, middle (white) can be manufactured by a simple and low cost process. It can also give a single LED of high brightness red, pink, orange, and white. Although red-green intermediate LEDs exist in the related art, the present invention can produce high-brightness red-green intermediate (yellow and orange) LEDs at a lower cost. The range of intermediate colors that can be produced by LEDs is widened. The manufacturing cost can be reduced as a single LED. By changing the type of substrate and the type of active layer, it is possible to manufacture LEDs that produce various white and intermediate colors. As it is an LED, it is small and light in low voltage and is easy to handle. Large demands can be expected for applications such as cropping and marking.

Claims (13)

And an epitaxial layer having a structure formed of a semiconductor crystal or an insulator crystal added with an impurity to be the center of fluorescence, and having a structure formed on the substrate to emit light by current injection. And a light emitting substrate LED device which emits any one of intermediate colors of reddish purple, pink, yellow, orange and white by mixing light emission of two different wavelengths with fluorescence obtained by photo-exciting a substrate by the light emission. . 2. The peak wavelength of the two types of light emission according to claim 1, wherein the element composition, the impurity element species, the impurity element amount, the substrate thickness of the crystal substrate, and the structure of the epitaxial layer are changed. A light emitting substrate LED device characterized in that the peak intensity ratio is adjusted to change the color tone of light emission of the device. The light emitting substrate LED device according to claim 2, wherein the substrate is an n-type AlGaAs substrate, and the active layer of the epitaxial layer is any one of a ZnSe layer, a ZnSSe / ZnCdSe layer, and a ZnSeTe layer. 4. The light emitting substrate LED device according to claim 3, wherein Si is used as an n-type impurity doped into the AlGaAs substrate. The light emitting substrate LED device according to claim 2, wherein the substrate is a GaP substrate and the active layer of the epitaxial layer is a GaInN layer. The light emitting substrate LED device according to claim 5, wherein Zn and O are used as impurities to dope the GaP substrate. The light emitting substrate LED device according to claim 2, wherein the substrate is a ZnSe substrate, and the active layer of the epitaxial layer is any one of a ZnSe layer, a ZnSe / ZnCdSe layer, and a ZnSeTe layer. 8. The light emitting substrate LED device according to claim 7, wherein Al and I are used as impurities to dope the ZnSe substrate. 8. The light emitting substrate LED device according to claim 7, wherein Al and I are used as impurities to dope the ZnSe substrate. 8. The light emitting substrate LED device according to claim 7, wherein Ag and I are used as impurities to dope the ZnSe substrate. The light emitting device according to claim 1, wherein the substrate is an n-type or p-type GaN single crystal substrate including oxygen atoms, carbon atoms, or nitrogen voids, and the active layer of the epitaxial layer is a Ga _1-X In _X N layer. Substrate LED device. 12. The light emitting structure of claim 11, wherein the light emitting structure has a multilayer structure including Ga _1-X In _X N, and the wavelength of light emitted from the light emitting structure is in the range of 400 nm to 510 nm, and the fluorescence emission from the GaN substrate is generated. A light emitting substrate LED device characterized in that the wavelength is 520nm to 650nm. The method according to claim 12, wherein the color tone of the obtained white light can be changed from the warm color system to the warm color system by adjusting the thickness of the GaN substrate in the range of 50 µm to 2 mm and changing the emission wavelength from the light emitting structure. Light emitting substrate LED device, characterized in that.